PERF: use isinstance(obj, Foo) instead of ABCFoo (pandas-dev#34040)

Author: jbrockmendel

pandas/core/frame.py

@@ -47,6 +47,7 @@
     Axis,
     Dtype,
     FilePathOrBuffer,
+    FrameOrSeriesUnion,
     IndexKeyFunc,
     Label,
     Level,
@@ -106,12 +107,6 @@
     needs_i8_conversion,
     pandas_dtype,
 )
-from pandas.core.dtypes.generic import (
-    ABCDataFrame,
-    ABCIndexClass,
-    ABCMultiIndex,
-    ABCSeries,
-)
 from pandas.core.dtypes.missing import isna, notna
 
 from pandas.core import algorithms, common as com, nanops, ops
@@ -1838,7 +1833,7 @@ def to_records(
                   dtype=[('I', 'S1'), ('A', '<i8'), ('B', '<f8')])
         """
         if index:
-            if isinstance(self.index, ABCMultiIndex):
+            if isinstance(self.index, MultiIndex):
                 # array of tuples to numpy cols. copy copy copy
                 ix_vals = list(map(np.array, zip(*self.index.values)))
             else:
@@ -1851,7 +1846,7 @@ def to_records(
             count = 0
             index_names = list(self.index.names)
 
-            if isinstance(self.index, ABCMultiIndex):
+            if isinstance(self.index, MultiIndex):
                 for i, n in enumerate(index_names):
                     if n is None:
                         index_names[i] = f"level_{count}"
@@ -2782,7 +2777,7 @@ def __getitem__(self, key):
             # The behavior is inconsistent. It returns a Series, except when
             # - the key itself is repeated (test on data.shape, #9519), or
             # - we have a MultiIndex on columns (test on self.columns, #21309)
-            if data.shape[1] == 1 and not isinstance(self.columns, ABCMultiIndex):
+            if data.shape[1] == 1 and not isinstance(self.columns, MultiIndex):
                 data = data[key]
 
         return data
@@ -3600,7 +3595,7 @@ def reindexer(value):
         elif isinstance(value, DataFrame):
             # align right-hand-side columns if self.columns
             # is multi-index and self[key] is a sub-frame
-            if isinstance(self.columns, ABCMultiIndex) and key in self.columns:
+            if isinstance(self.columns, MultiIndex) and key in self.columns:
                 loc = self.columns.get_loc(key)
                 if isinstance(loc, (slice, Series, np.ndarray, Index)):
                     cols = maybe_droplevels(self.columns[loc], key)
@@ -3649,7 +3644,7 @@ def reindexer(value):
 
         # broadcast across multiple columns if necessary
         if broadcast and key in self.columns and value.ndim == 1:
-            if not self.columns.is_unique or isinstance(self.columns, ABCMultiIndex):
+            if not self.columns.is_unique or isinstance(self.columns, MultiIndex):
                 existing_piece = self[key]
                 if isinstance(existing_piece, DataFrame):
                     value = np.tile(value, (len(existing_piece.columns), 1))
@@ -4337,9 +4332,7 @@ def set_index(
 
         missing: List[Label] = []
         for col in keys:
-            if isinstance(
-                col, (ABCIndexClass, ABCSeries, np.ndarray, list, abc.Iterator)
-            ):
+            if isinstance(col, (Index, Series, np.ndarray, list, abc.Iterator)):
                 # arrays are fine as long as they are one-dimensional
                 # iterators get converted to list below
                 if getattr(col, "ndim", 1) != 1:
@@ -4368,19 +4361,19 @@ def set_index(
         names = []
         if append:
             names = list(self.index.names)
-            if isinstance(self.index, ABCMultiIndex):
+            if isinstance(self.index, MultiIndex):
                 for i in range(self.index.nlevels):
                     arrays.append(self.index._get_level_values(i))
             else:
                 arrays.append(self.index)
 
         to_remove: List[Label] = []
         for col in keys:
-            if isinstance(col, ABCMultiIndex):
+            if isinstance(col, MultiIndex):
                 for n in range(col.nlevels):
                     arrays.append(col._get_level_values(n))
                 names.extend(col.names)
-            elif isinstance(col, (ABCIndexClass, ABCSeries)):
+            elif isinstance(col, (Index, Series)):
                 # if Index then not MultiIndex (treated above)
                 arrays.append(col)
                 names.append(col.name)
@@ -4625,7 +4618,7 @@ def _maybe_casted_values(index, labels=None):
 
         if not drop:
             to_insert: Iterable[Tuple[Any, Optional[Any]]]
-            if isinstance(self.index, ABCMultiIndex):
+            if isinstance(self.index, MultiIndex):
                 names = [
                     (n if n is not None else f"level_{i}")
                     for i, n in enumerate(self.index.names)
@@ -4636,7 +4629,7 @@ def _maybe_casted_values(index, labels=None):
                 names = [default] if self.index.name is None else [self.index.name]
                 to_insert = ((self.index, None),)
 
-            multi_col = isinstance(self.columns, ABCMultiIndex)
+            multi_col = isinstance(self.columns, MultiIndex)
             for i, (lev, lab) in reversed(list(enumerate(to_insert))):
                 if not (level is None or i in level):
                     continue
@@ -5240,7 +5233,7 @@ def sort_index(
                 level, ascending=ascending, sort_remaining=sort_remaining
             )
 
-        elif isinstance(labels, ABCMultiIndex):
+        elif isinstance(labels, MultiIndex):
             from pandas.core.sorting import lexsort_indexer
 
             indexer = lexsort_indexer(
@@ -5607,14 +5600,14 @@ def swaplevel(self, i=-2, j=-1, axis=0) -> "DataFrame":
 
         axis = self._get_axis_number(axis)
 
-        if not isinstance(result._get_axis(axis), ABCMultiIndex):  # pragma: no cover
+        if not isinstance(result._get_axis(axis), MultiIndex):  # pragma: no cover
             raise TypeError("Can only swap levels on a hierarchical axis.")
 
         if axis == 0:
-            assert isinstance(result.index, ABCMultiIndex)
+            assert isinstance(result.index, MultiIndex)
             result.index = result.index.swaplevel(i, j)
         else:
-            assert isinstance(result.columns, ABCMultiIndex)
+            assert isinstance(result.columns, MultiIndex)
             result.columns = result.columns.swaplevel(i, j)
         return result
 
@@ -5635,16 +5628,16 @@ def reorder_levels(self, order, axis=0) -> "DataFrame":
         DataFrame
         """
         axis = self._get_axis_number(axis)
-        if not isinstance(self._get_axis(axis), ABCMultiIndex):  # pragma: no cover
+        if not isinstance(self._get_axis(axis), MultiIndex):  # pragma: no cover
             raise TypeError("Can only reorder levels on a hierarchical axis.")
 
         result = self.copy()
 
         if axis == 0:
-            assert isinstance(result.index, ABCMultiIndex)
+            assert isinstance(result.index, MultiIndex)
             result.index = result.index.reorder_levels(order)
         else:
-            assert isinstance(result.columns, ABCMultiIndex)
+            assert isinstance(result.columns, MultiIndex)
             result.columns = result.columns.reorder_levels(order)
         return result
 
@@ -5913,7 +5906,8 @@ def extract_values(arr):
             # Does two things:
             # 1. maybe gets the values from the Series / Index
             # 2. convert datelike to i8
-            if isinstance(arr, (ABCIndexClass, ABCSeries)):
+            # TODO: extract_array?
+            if isinstance(arr, (Index, Series)):
                 arr = arr._values
 
             if needs_i8_conversion(arr.dtype):
@@ -5925,7 +5919,8 @@ def extract_values(arr):
 
         def combiner(x, y):
             mask = isna(x)
-            if isinstance(mask, (ABCIndexClass, ABCSeries)):
+            # TODO: extract_array?
+            if isinstance(mask, (Index, Series)):
                 mask = mask._values
 
             x_values = extract_values(x)
@@ -7020,8 +7015,8 @@ def _gotitem(
         self,
         key: Union[str, List[str]],
         ndim: int,
-        subset: Optional[Union[Series, ABCDataFrame]] = None,
-    ) -> Union[Series, ABCDataFrame]:
+        subset: Optional[FrameOrSeriesUnion] = None,
+    ) -> FrameOrSeriesUnion:
         """
         Sub-classes to define. Return a sliced object.
 
@@ -8225,7 +8220,7 @@ def _count_level(self, level, axis=0, numeric_only=False):
         count_axis = frame._get_axis(axis)
         agg_axis = frame._get_agg_axis(axis)
 
-        if not isinstance(count_axis, ABCMultiIndex):
+        if not isinstance(count_axis, MultiIndex):
             raise TypeError(
                 f"Can only count levels on hierarchical {self._get_axis_name(axis)}."
             )

pandas/core/reshape/melt.py

@@ -7,13 +7,12 @@
 
 from pandas.core.dtypes.common import is_extension_array_dtype, is_list_like
 from pandas.core.dtypes.concat import concat_compat
-from pandas.core.dtypes.generic import ABCMultiIndex
 from pandas.core.dtypes.missing import notna
 
 from pandas.core.arrays import Categorical
 import pandas.core.common as com
 from pandas.core.frame import DataFrame, _shared_docs
-from pandas.core.indexes.base import Index
+from pandas.core.indexes.api import Index, MultiIndex
 from pandas.core.reshape.concat import concat
 from pandas.core.tools.numeric import to_numeric
 
@@ -33,15 +32,15 @@ def melt(
     # TODO: what about the existing index?
     # If multiindex, gather names of columns on all level for checking presence
     # of `id_vars` and `value_vars`
-    if isinstance(frame.columns, ABCMultiIndex):
+    if isinstance(frame.columns, MultiIndex):
         cols = [x for c in frame.columns for x in c]
     else:
         cols = list(frame.columns)
 
     if id_vars is not None:
         if not is_list_like(id_vars):
             id_vars = [id_vars]
-        elif isinstance(frame.columns, ABCMultiIndex) and not isinstance(id_vars, list):
+        elif isinstance(frame.columns, MultiIndex) and not isinstance(id_vars, list):
             raise ValueError(
                 "id_vars must be a list of tuples when columns are a MultiIndex"
             )
@@ -60,9 +59,7 @@ def melt(
     if value_vars is not None:
         if not is_list_like(value_vars):
             value_vars = [value_vars]
-        elif isinstance(frame.columns, ABCMultiIndex) and not isinstance(
-            value_vars, list
-        ):
+        elif isinstance(frame.columns, MultiIndex) and not isinstance(value_vars, list):
             raise ValueError(
                 "value_vars must be a list of tuples when columns are a MultiIndex"
             )
@@ -84,7 +81,7 @@ def melt(
         frame.columns = frame.columns.get_level_values(col_level)
 
     if var_name is None:
-        if isinstance(frame.columns, ABCMultiIndex):
+        if isinstance(frame.columns, MultiIndex):
             if len(frame.columns.names) == len(set(frame.columns.names)):
                 var_name = frame.columns.names
             else:

pandas/core/series.py

@@ -45,13 +45,7 @@
     is_object_dtype,
     is_scalar,
 )
-from pandas.core.dtypes.generic import (
-    ABCDataFrame,
-    ABCDatetimeIndex,
-    ABCMultiIndex,
-    ABCPeriodIndex,
-    ABCSeries,
-)
+from pandas.core.dtypes.generic import ABCDataFrame
 from pandas.core.dtypes.inference import is_hashable
 from pandas.core.dtypes.missing import (
     isna,
@@ -272,7 +266,7 @@ def __init__(
                         "Cannot construct a Series from an ndarray with "
                         "compound dtype.  Use DataFrame instead."
                     )
-            elif isinstance(data, ABCSeries):
+            elif isinstance(data, Series):
                 if index is None:
                     index = data.index
                 else:
@@ -3549,7 +3543,7 @@ def swaplevel(self, i=-2, j=-1, copy=True) -> "Series":
         Series
             Series with levels swapped in MultiIndex.
         """
-        assert isinstance(self.index, ABCMultiIndex)
+        assert isinstance(self.index, MultiIndex)
         new_index = self.index.swaplevel(i, j)
         return self._constructor(self._values, index=new_index, copy=copy).__finalize__(
             self, method="swaplevel"
@@ -3574,7 +3568,7 @@ def reorder_levels(self, order) -> "Series":
             raise Exception("Can only reorder levels on a hierarchical axis.")
 
         result = self.copy()
-        assert isinstance(result.index, ABCMultiIndex)
+        assert isinstance(result.index, MultiIndex)
         result.index = result.index.reorder_levels(order)
         return result
 
@@ -4661,8 +4655,8 @@ def to_timestamp(self, freq=None, how="start", copy=True) -> "Series":
         if copy:
             new_values = new_values.copy()
 
-        assert isinstance(self.index, (ABCDatetimeIndex, ABCPeriodIndex))
-        new_index = self.index.to_timestamp(freq=freq, how=how)
+        assert isinstance(self.index, PeriodIndex)
+        new_index = self.index.to_timestamp(freq=freq, how=how)  # type: ignore
         return self._constructor(new_values, index=new_index).__finalize__(
             self, method="to_timestamp"
         )
@@ -4688,8 +4682,8 @@ def to_period(self, freq=None, copy=True) -> "Series":
         if copy:
             new_values = new_values.copy()
 
-        assert isinstance(self.index, ABCDatetimeIndex)
-        new_index = self.index.to_period(freq=freq)
+        assert isinstance(self.index, DatetimeIndex)
+        new_index = self.index.to_period(freq=freq)  # type: ignore
         return self._constructor(new_values, index=new_index).__finalize__(
             self, method="to_period"
         )

pandas/io/formats/excel.py

@@ -14,9 +14,9 @@
 
 from pandas.core.dtypes import missing
 from pandas.core.dtypes.common import is_float, is_scalar
-from pandas.core.dtypes.generic import ABCIndex, ABCMultiIndex, ABCPeriodIndex
+from pandas.core.dtypes.generic import ABCIndex
 
-from pandas import Index
+from pandas import Index, MultiIndex, PeriodIndex
 import pandas.core.common as com
 
 from pandas.io.common import stringify_path
@@ -465,7 +465,7 @@ def _format_header_mi(self):
         coloffset = 0
         lnum = 0
 
-        if self.index and isinstance(self.df.index, ABCMultiIndex):
+        if self.index and isinstance(self.df.index, MultiIndex):
             coloffset = len(self.df.index[0]) - 1
 
         if self.merge_cells:
@@ -507,7 +507,7 @@ def _format_header_regular(self):
 
             if self.index:
                 coloffset = 1
-                if isinstance(self.df.index, ABCMultiIndex):
+                if isinstance(self.df.index, MultiIndex):
                     coloffset = len(self.df.index[0])
 
             colnames = self.columns
@@ -526,7 +526,7 @@ def _format_header_regular(self):
                 )
 
     def _format_header(self):
-        if isinstance(self.columns, ABCMultiIndex):
+        if isinstance(self.columns, MultiIndex):
             gen = self._format_header_mi()
         else:
             gen = self._format_header_regular()
@@ -545,7 +545,7 @@ def _format_header(self):
         return itertools.chain(gen, gen2)
 
     def _format_body(self):
-        if isinstance(self.df.index, ABCMultiIndex):
+        if isinstance(self.df.index, MultiIndex):
             return self._format_hierarchical_rows()
         else:
             return self._format_regular_rows()
@@ -569,15 +569,15 @@ def _format_regular_rows(self):
             else:
                 index_label = self.df.index.names[0]
 
-            if isinstance(self.columns, ABCMultiIndex):
+            if isinstance(self.columns, MultiIndex):
                 self.rowcounter += 1
 
             if index_label and self.header is not False:
                 yield ExcelCell(self.rowcounter - 1, 0, index_label, self.header_style)
 
             # write index_values
             index_values = self.df.index
-            if isinstance(self.df.index, ABCPeriodIndex):
+            if isinstance(self.df.index, PeriodIndex):
                 index_values = self.df.index.to_timestamp()
 
             for idx, idxval in enumerate(index_values):
@@ -609,7 +609,7 @@ def _format_hierarchical_rows(self):
             # with index names (blank if None) for
             # unambiguous round-trip, unless not merging,
             # in which case the names all go on one row Issue #11328
-            if isinstance(self.columns, ABCMultiIndex) and self.merge_cells:
+            if isinstance(self.columns, MultiIndex) and self.merge_cells:
                 self.rowcounter += 1
 
             # if index labels are not empty go ahead and dump